Who talks more…men or women?

There is nothing I love more than a clever phrase to describe a phenomena that bothers me.  Last night I found such a phrase in a Jezebel article about gender differences in number of words spoken per day.

The phrase: public domain data.
This describes the phenomena of data that’s been so often cited that no one seems to think you need a reference for it anymore (think “you need 8 glasses of water per day”).  
In this particular article, she brings it up in the context of the assertion that women say more words per day than men do.  The most commonly cited data puts it at anywhere from double to quadruple the number…quite a difference by any stretch of the imagination.  The problem is the numbers are apparently quite fictional.
The Jezebel piece links to this piece on a blog from UPENN where someone actually tried to track down a study that ever established the numbers quoted.  They found that it apparently originated with James Dobson sometime in the late 80’s or early 90’s, and that he did not seem to have a study to back him up (his actual quote was that “God gave women 50,000 words for the day, and men 25,000”).
From there it seems to have changed and distorted, but it does not seem anyone has actually sat down and counted the number of words men and women said.
If you think about it, this would be a really hard study to do.  I mean, I think most people would knee-jerk agree that women talk more than men, but we have to remember that we’re only thinking of social situations. Your choice of profession would HEAVILY skew how many words you were saying per day.  I mean, my brother the biology teacher is by far the quietest member of my immediate family when we get together.  However, I would hazard a guess that 5 days out of the week he likely says more words than I do.  I mean, he teaches.  He has to say words.  I spend at least half my day analyzing data.  No words needed.  I also have a longer commute (no talking needed) and unless I’m giving a presentation all of my meetings involve giving people equal time to talk.  But of course people who met us wouldn’t count his professional speaking. It would be clear to everyone that I talk more…unless you had a researcher study us on an average day.
Additionally, even if this stat had been true in 1993, how would we count it today?  In 1993, most social communication was verbal.  How would we count blogging, Twitter, Reddit or texting? It feels strange to count those as words, but also strange to not include them.  
I mean, the little lord’s been taking his morning nap for the past hour or so.  In that time I have commented on 4 blogs, written two emails, one Facebook message, and written this post.  For research purposes though, I haven’t said a word.  Interesting, isn’t it?

3 thoughts on “Who talks more…men or women?

  1. I tend to take half the conversational space, no matter how many people are present. Not likely to be actually true, but close. Yet there are situations in which I say almost nothing, and others where I avoid having to go because I would be expected to talk and I don't want to.

    Plus I talk to myself, sort of out loud, a lot. Do those count? And as you note, texts, tweets, comments…how do you score a forwarded link or a forwarded photo?

    Eh, why bother? Everyone knows women talk more than men, right?


  2. Well, if you wanted to measure how many words a person speaks, you'd need to clip a microphone to their shirt. (And somehow convince them to behave normally.)

    Then you'd have to ask whether a person talking to themselves counts. And you'd have to distinguish “talking to self” from “giving a presentation” and “teaching a class.”

    If there really is a difference…

    Great point, though, on “public domain data”.


  3. I think when I read the comments by James Dobson, I was also wondering if talking to babies counts. I mean, I talk to the little lord most of the time I'm with him, but that's more about trying to help engage him and language development than expression.

    I think what this gets down to is that there's lots of different types of speech, and we'd have to be quite precise with what we were measuring before we figured out who said more.


Comments are closed.